The official repo of MiniMax-Text-01 and MiniMax-VL-01, large-language-model & vision-language-model based on Linear Attention
- large-language-models
- llm
- minimax-text-01
- minimax-vl-01
132,241 visible repositories in the index
25–48 of 161 · Most stars
The official repo of MiniMax-Text-01 and MiniMax-VL-01, large-language-model & vision-language-model based on Linear Attention
Make videos on the computer you already own. FreeVideo runs MiniMax H3 in as little as 8 GB of VRAM and 16 GB of RAM, and adapts its acceleration path to your hardware.
Python package to easily retrain OpenAI's GPT-2 text-generating model on new texts
基于LangChain和ChatGLM-6B等系列LLM的针对本地知识库的自动问答
A modern selfhosted media management system for your media library
An autonomous company working 24/7 on your own PC - Windows/Linux/macOS.
MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model.
A minimal yet professional single agent demo project that showcases the core execution pipeline and production-grade features of agents.
TokenSpeed is a speed-of-light LLM inference engine.
Implementation of papers in 100 lines of code.
The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.
Survival analysis in Python
Bitcoin arbitrage - opportunity detector
Battle-tested agent skills mapping the footguns of Kotlin, Compose Multiplatform and the desktop JVM — mined from a production music app, not from documentation.
A simple, performant, and scalable Jax LLM!
TexTools is a UV and Texture toolset created several years ago for Blender and Max by @renderhjs. In this open repository, originally created by @SavMartin, we develop the current version of this popular Blender add-on.
A local dashboard for tracking your Claude Code token usage, costs, and session history. Pro and Max subscribers get a progress bar. This gives you the full picture.
label-smooth, amsoftmax, partial-fc, focal-loss, triplet-loss, lovasz-softmax. Maybe useful
Chinese NER(Named Entity Recognition) using BERT(Softmax, CRF, Span)
TranslatedManju Laoli (漫剧老李, "Comic Drama Old Li") AIGC full-workflow creation Skill · V7.4.26 multi-skill suite edition (Short-Drama Director Suite): 1 master controller + 12 sub-skills · four-layer deep architecture — genre routing / material scouting / script / assets / storyboard / three major fight-scene engines / feeding / quality inspection across the whole chain. Six pre-locks + I art-style lock · 50 rule libraries · machine checks C1~C33 | H1~H12 + six feeding gates + P4.5 storyboard machine checks (P450~P458) · material forms M0/M1/M2 (this package does not hold images) · outputs seven-section prompt sets that can be fed directly into Seedance 2.5 / 2.0 / MiniMax H3. Beta version, a secondary modification that integrates open-source content.
Keras + Hyperopt: A very simple wrapper for convenient hyperparameter optimization
AI-native agent harness for coding workflows by python: multi-model LLM orchestration, stateful sessions, tool governance, traceable delivery, and provider routing for GPT, Claude, DeepSeek, Qwen, Kimi, GLM, and MiniMax.
Retarget human motion from video to an Adobe Mixamo-rigged 3D character.
Automatically "block" people in images (like Black Mirror) using a pretrained neural network.